China is executing a sophisticated new geopolitical strategy dubbed “token diplomacy,” leveraging its rapidly advancing open-source artificial intelligence models to build influence and technological reliance across the Global South. Semafor reports that by offering powerful AI tools at a fraction of the cost of American alternatives, Beijing aims to shape global tech standards and establish itself as the indispensable partner for developing nations in the AI era, a digital evolution of the Belt and Road Initiative that could prove far more consequential than its infrastructure predecessor.
This strategy was on full display at the recent UN AI for Good summit in Geneva. While prominent U.S. frontier lab leaders and Trump administration officials were notably absent, a massive delegation of Chinese government officials and executives fanned out to pitch their vision. Their message to ministers from countries like Pakistan, Zambia, and across the Global South was clear: free, Chinese-made open-source models are the future of their economies. The contrast with American absence was stark and deliberate — and it did not go unnoticed by the ministers in the room.
The practical appeal of China’s offer is straightforward. A government ministry in a developing nation that wants to build an AI-powered public services platform faces a stark choice: pay premium prices for access to U.S. models that can be restricted or revoked by executive order, or download a Chinese open-weight model that can be run locally, modified freely, and never switched off by a foreign government.
For nations that have experienced the disruption caused by U.S. export controls, including the sudden unavailability of Anthropic’s Mythos and Fable models for non-U.S. entities, the appeal of technological sovereignty through Chinese open source is not ideological but pragmatic.
Building the AI Belt and Road
China’s token diplomacy is a direct digital evolution of its Belt and Road Initiative. Instead of exporting ports, railways, and telecom networks, Beijing is now exporting AI capabilities. Wang Jian, a prominent Chinese computer scientist and former Microsoft Asia executive who now runs a government-backed AI research institute, articulated this vision in Geneva: China’s AI can be a “resource” for other countries much like energy. “Having a choice for the rest of the world is very important,” he said.
This approach capitalizes on a growing desire among developing nations for technological independence from the United States — a sentiment significantly exacerbated when the U.S. abruptly imposed export controls on Anthropic’s Mythos and Fable models, causing massive disruptions for non-U.S. entities reliant on them.
As Xi’s WAIC open-source gambit revealed China’s decade-long AI strategy, the offer of open-source models that can be downloaded and modified locally presents an attractive, seemingly restriction-free alternative to American platforms that can be switched off by executive order.
Shaping Global Standards
The ultimate goal of this diplomacy extends beyond market share; it is about setting the rules of the road for the next era of technology. Chinese tech firms have already successfully shaped UN standards for facial recognition and automated driving systems. By embedding their AI models into the digital infrastructure of the Global South, Beijing ensures that future technological ecosystems will be compatible with Chinese standards rather than American ones. As seen in a previous analysis of China’s WAICO push, this is a governance offensive, not a goodwill gesture.
Chinese leader Xi Jinping reinforced this mission at the World AI Conference in Shanghai, urging nations to “jointly oppose overstretching the national security concept in the field of AI” and warning that “AI development should not be a solo performance by a single country.” He pledged to help align global rules through the newly formed World AI Cooperation Organization (WAICO), a 29-country alliance that notably excludes the U.S.
However, the strategy faces real headwinds. Many governments remain wary of building on Chinese open-source models that haven’t disclosed their training data and carry a Beijing-aligned framing on sensitive topics. Nkundwe Mwasaga, director general of the ICT Commission in Tanzania, a country whose national fiber network was built by Chinese firms and financed by Chinese state loans, acknowledged that America “leads the pack,” while China is “just catching up.” His country is developing a sovereign AI model that local startups can build using either Chinese or American AI models, illustrating the pragmatic hedging strategy many developing nations are adopting.
The United States, for its part, has been slow to articulate a coherent counter-strategy for the Global South. The Trump administration’s focus on domestic AI buildout and export controls has left a diplomatic vacuum that China is aggressively filling. American AI companies, preoccupied with the domestic market and the competitive pressure from Chinese models, have not mounted a concerted effort to engage developing nations on AI infrastructure.
If Washington does not develop a credible alternative to China’s token diplomacy before the September summit, it may find that the geopolitical map of AI has already been redrawn in Beijing’s favor. The battle for the Global South’s AI future is far from decided — but China is currently the only major power showing up to fight for it.